Full factorial analysis of mammalian and avian influenza polymerase subunits suggests a role of an efficient polymerase for virus adaptation

Abstract

Amongst all the internal gene segments (PB2. PB1, PA, NP, M and NS), the avian PB1 segment is the only one which was reassorted into the human H2N2 and H3N2 pandemic strains. This suggests that the reassortment of polymerase subunit genes between mammalian and avian influenza viruses might play roles for interspecies transmission. To test this hypothesis, we tested the compatibility between PB2, PB1, PA and NP derived from a H5N1 virus and a mammalian H1N1 virus. All 16 possible combinations of avian-mammalian chimeric viral ribonucleoproteins (vRNPs) were characterized. We showed that recombinant vRNPs with a mammalian PB2 and an avian PB1 had the strongest polymerase activities in human cells at all studied temperature. In addition, viruses with this specific PB2-PB1 combination could grow efficiently in cell cultures, especially at a high incubation temperature. These viruses were potent inducers of proinflammatory cytokines and chemokines in primary human macrophages and pneumocytes. Viruses with this specific PB2-PB1 combination were also found to be more capable to generate adaptive mutations under a new selection pressure. These results suggested that the viral polymerase activity might be relevant for the genesis of influenza viruses of human health concern.

Figure 1. Characterization of recombinant vRNPs generated from transfected cells.

The origins of PB2, PB1, PA and NP in each recombinant vRNP were as shown (A = avian, M = mammalian). A) Luciferase reporter assay for influenza viral polymerase activity. Polymerase activities (mean±SE) of recombinant vRNPs in 293T cells incubated at 32°C (top panel), 37°C (middle panel) and 40°C (bottom panel). All data were determined from three independent experiments. The polymerase activities of WSN were set as 100% as references. B) Detection of NA mRNA, cRNA, and vRNA by primer extension assays. Signals for the mRNA, cRNA and vRNA were as shown. C) Immunoprecipitation of TAP-tagged PA. Nuclear lysates expressing different combinations of chimeric viral polymerase complexes were immunoprecipitated by immunoglobinlin G-Sepharose. The amounts of PB2 and Pol II_o coimmunoprecipitated with TAP-PA were determined by Western blot techniques.

Figure 2. Polymerase activity of vRNPs with chimeric PB1.

Various combinations of chimeric PB1 (Mutants 1 to 4, right panel) derived from mammalian (hatched bar) and avian (solid bar) PB1 were tested in the luciferase reporter assay in 293T cells incubated at 37°C. The length of PB1 fragment in each region was indicated. The effect of these chimeric PB1 on vRNP in a WSN (left top panel) or Indo5 (left bottom panel) background were shown. The activities of the wild-type control in the corresponding backgrounds were set as 100% for references.

Figure 3. Luciferase reporter assay for influenza viral polymerase activity.

Polymerase activities (mean±SE) of recombinant vRNPs generated in CEF cells at 37°C are as shown.

Figure 4. Characterization of recombinant viruses with chimeric polymerase complexes.

The origins of PB2, PB1, PA and NP in each recombinant virus was as shown (A = avian, M = mammalian). A) Plaque size (mean±SE) of the wild type (MMMM) and recombinant viruses in MDCK cells at 72 hours post-infection. B) Growth properties of the WSN (MMMM) and recombinant viruses in MDCK cells. The number of infectious progeny viral particles generated from MDCK cells infected with the corresponding virus at a MOI of 0.01 was determined by standard plaque assay. Mutant AMAA and AMMM were significantly attenuated (ANOVA, p<0.05). *At 8 hours post-infection, the amounts of infectious progeny of MAAA and MAMM were significantly higher than the wild type control (t-test, p<0.05). C) NA-specific primer extension assays. Total RNA from MDCK cells infected with influenza virus at a MOI of 2 was harvested at 6 hours post-infection. Signals for the mRNA, cRNA and vRNA were as shown. D) Western blot analysis of influenza PA and cellular Pol II from infected MDCK cells. Total cell lysates from cells infected at 2 MOI were harvested at 6 hours post-infection. Signals for PA and ubiquitinated (Ubi-Pol II), hyperphosphorylated (II_o) and hypophosphorylated (II_a) Pol II were indicated. β-actin was used as a control. The PA level of the MAMM mutant was more abundant than the wild type (MMMM), a further illustration of the faster transcription kinetics of this chimeic vRNP. The experiment was repeated three times with comparable results.

Figure 5. Cytokine and chemokine gene expression profiles (mean±SE) from primary human macrophages (A) and pneuomocytes (B).

Total RNA from cells infected at a MOI of 2 was harvested at the indicated time points and tested by the corresponding quantitative RT-PCR assays as indicated. The data were the averages of triplicate assays. The recombinant viruses used in the experiments were as shown.

Figure 6. Serial passage of the MAMM mutant in the presence of oseltamivir.

A) Viral titers in culture supernatants. The wild type and MAMM mutant were serially passaged in MDCK cells in the presence of oseltamivir. The viral cultures were harvested at 72 hr post-infection and the supernatant from passages 1, 3, 5 and 7 were titrated by standard plaque assays. A representative data set from duplicated experiments was shown. B) Protein sequence of the NA stalk. The passage histories and NA sequences from the parental controls, passaged controls and MAMM escape mutants were as shown. The IC₅₀ of oseltamivir towards these viruses were presented.